6 People, 7 PPT Slides, Secured 800 Million in Funding: Will AI Become the Next Internet Bubble?
-
Recently, Zhu Xiaohu, a partner at GSR Ventures, stated in an interview, 'I believe in AGI, but I believe more in what can be commercialized immediately.'
Silicon Valley investors think differently; they prefer 'taking a shot in the dark.'
As a result, the funding rounds for AI startups have become increasingly surreal, with valuations reaching absurd heights:
The latest valuation for the AI programmer Devin team may reach $2 billion; Scale AI's valuation is set to reach $13 billion;
Amazon has completed its full $4 billion investment in AI company Anthropic;
Heygen has secured $60 million in funding with a valuation of $500 million;
Mistral AI, just four weeks old with only 6 employees and a 7-page pitch deck, has raised $800 million; In this AI frenzy, some seem to detect echoes of the late 20th-century internet bubble burst and begin to question whether AI might be another bubble...
Currently, the AI market is permeated with intense FOMO (Fear of Missing Out) sentiment. Investors, afraid of missing opportunities, are pouring massive funds into the AI field.
It's reported that in 2023's 691 deals, investors injected nearly $30 billion into the AI industry. AI startups alone accounted for one-third of all investments in 2023. In 2024, investors' enthusiasm for AI continues to grow. Throughout March, news of financing in the AI sector has been frequently reported.
On March 31, Cognition Labs, a US-based AI-assisted programming startup founded by a Silicon Valley Chinese team, is seeking a new round of funding at a valuation of $2 billion. Silicon Valley venture capital firms such as Founders Fund are in talks to invest in this round.
It is worth noting that the company was established less than half a year ago and only launched its code tool Devin, touted as the "first AI programmer close to human capability," in early March this year. So far, it has not generated any actual revenue. This zero-revenue AI startup announced $21 million in funding at a $350 million valuation in March. In less than a month, Cognition Labs' projected valuation has surged over fivefold.
On March 30, venture firm Accel was negotiating to lead a new funding round for data labeling company Scale AI, potentially boosting its valuation by nearly 80% to approximately $13 billion.
On March 28, AI startup Cohere entered discussions for a new funding round exceeding $500 million, targeting a valuation as high as $5 billion. On the same day, Amazon announced an additional investment of $2.75 billion in AI startup Anthropic, bringing its total investment to $4 billion, the company's largest external investment to date.
Also on the same day, HeyGen, a three-year-old AI video generation startup, is in the process of raising $60 million in new funding, with its latest valuation reaching $440 million, six times (600%) its valuation from four months ago.
In early March, AI search startup Perplexity secured funding, with its valuation rising from $520 million a few months ago to $1 billion. This isn't even the most magical part. The real shocker is French startup Mistral AI, hailed as the 'European OpenAI'.
Founded in Paris in May 2023, Mistral AI started with just 6 employees and no tangible product. Yet, within merely 4 weeks, the company secured €105 million (approximately ¥827 million) in seed funding using nothing but a 7-slide pitch deck.
By December of the same year, Mistral AI raised an additional €385 million (about $415 million) in Series A funding, skyrocketing its valuation to $2 billion. Don't be fooled by these companies' skyrocketing valuations - their actual revenue hasn't improved much.
According to a Sequoia Capital investor's speech in March, the AI industry spent $50 billion last year just on Nvidia chips for training large models, while generating only $3 billion in revenue. This means excluding electricity costs, the AI industry's expenses are about 17 times its income.
Beyond low profitability and funding struggles to keep up with cash burn, these AI startups haven't established a viable commercialization path either. Therefore, many industry insiders have begun to question: Will AI become the next internet bubble?
In fact, some AI startups are already "weathering the storm."
For example, the UK's popular AI unicorn Stability AI. At the end of last month, Stability AI suddenly announced that CEO Emad Mostaque had resigned and stepped down from the company's board of directors.
Prior to this, the company's core Stable Diffusion research team had collectively resigned, including the research team leader and primary authors of the seminal paper.
Founded in 2020, Stability AI gained nearly $100 million in funding through its Stable Diffusion technology, reaching a $1 billion valuation and becoming a new unicorn company. However, as of October 2023, the company's burn rate was approximately $8 million per month. Combined with the absence of a suitable profit model, Stability AI depleted its financial resources within a year and even incurred millions of dollars in debts to Amazon.
Subsequently, Stability AI attempted to raise more funds at a $4 billion valuation but was unsuccessful.
Not only Stability AI is facing management changes among AI startups, but even the core team of the US AI startup star Inflection AI, valued at $4 billion, has departed. The CEO and Chief Scientist of Inflection AI have joined Microsoft, where they are leading a newly established AI division and have taken a large number of Inflection AI employees with them.
This wave of AI has been surging for over a year, causing fluctuations in global stock markets. There is ongoing debate about whether AI will develop into the biggest bubble in history.
Demis Hassabis, co-founder of DeepMind, believes that billions of dollars invested in generative AI startups and products have led to significant hype, similar to what was seen with cryptocurrencies, overshadowing scientific research. However, some hold an optimistic view, considering the current development of AI technology as 'a true revolution'.